Quantum computers: which startup is ahead?
In our quantum computing market deck, you will find everything you need to understand the market
SUMMARY
PsiQuantum is ahead overall in the private quantum-computing race, with QuEra and Atom Computing as its closest technical challengers and Pasqal leading the commercial market that exists today.
PsiQuantum’s lead comes from the completeness of its plan rather than from a working public machine. No other private startup has combined comparable funding, dedicated computing sites, semiconductor production work, component development and system-level government review.
That lead is still fragile. PsiQuantum has not shown a complete computer running customer workloads, while QuEra has already produced the strongest public evidence of logical qubits, repeated error correction and fault-tolerant operations on working hardware.
Atom Computing is the most credible bridge between large physical scale and logical computation. Its AC1000 platform contains more than 1,200 physical qubits, but its strongest error-correction demonstrations have so far used smaller parts of earlier systems.
Pasqal is winning a different race. It has disclosed revenue, delivered complete processors to recognizable customers and built a repeat-installation record that none of the more ambitious fault-tolerant startups currently matches.
Neutral atoms are the strongest architecture in the startup field today. QuEra leads the research, Atom Computing has packaged the largest gate-based platform, Pasqal has commercialized the technology and planqc has won contracts for larger future systems.
Funding is a useful indicator of industrial ambition, but it is a poor substitute for technical proof. PsiQuantum has raised far more than its rivals, yet QuEra’s peer-reviewed experiments currently reveal more about how error-corrected quantum computation behaves inside a real machine.
The hardest lead to copy may belong to PsiQuantum because its moat spans chips, detectors, optical switches, packaging, fibre connections, cryogenic engineering and foundry relationships. The same breadth creates a nasty integration risk: one weak component can hold back the whole computer.
Silicon startups may eventually gain the cleanest manufacturing route. Quantum Motion and Diraq are already using standard 300 mm semiconductor processes, although their operating processors remain far smaller than the leading neutral-atom machines.
Government backing points most strongly toward PsiQuantum, while customer evidence points toward Pasqal. AWS gives QuEra the best public-cloud channel, Microsoft gives Atom Computing a powerful system partner, and Photonic has a serious investor and technical collaborator in the same company.
No startup has yet shown a repeatable economic advantage over the best classical computers. The current ranking therefore measures who has built the most credible route to usefulness, not who has already delivered a useful fault-tolerant quantum computer.
PsiQuantum stays first because it has the strongest complete industrial program. QuEra can overtake it by delivering Libra near its stated specifications, while Atom Computing could move into the top two by running repeated error-corrected algorithms across its full packaged platform.
This market map, featured in our quantum computing market deck, highlights top companies and startups in the quantum computing market
Quantum computers: which startup is ahead?
Which startups are we actually comparing?
We currently see thirteen private quantum-computing startups worth comparing, but only six have a convincing claim to the lead.
We include companies trying to build a complete quantum computer, rather than businesses focused only on software, control electronics, sensors, networking or cybersecurity. We also limit the comparison to independent private companies.
That removes several famous names. Quantinuum, IQM, Xanadu and Infleqtion have entered the public markets. IonQ, Rigetti, D-Wave and Quantum Computing Inc. were already listed. IonQ also acquired Oxford Ionics. IBM, Google, Microsoft and Intel remain major competitors, but quantum computing represents only one part of their businesses.
PsiQuantum and Photonic use light. QuEra, Atom Computing, Pasqal and planqc use neutral atoms. Quantum Motion, Diraq and Silicon Quantum Computing use silicon qubits. OQC builds superconducting machines, while Alice & Bob and Nord Quantique are developing superconducting qubits that may require less error-correction hardware. Quandela builds photonic systems already available through the cloud or at customer sites.
Funding totals require caution because equity, government grants and proposed incentives are not equivalent. We count confirmed private funding where possible and label other support separately.
| Startup | What it is building | Funding raised or publicly committed |
|---|---|---|
| PsiQuantum | Large fault-tolerant computers using photonic chips | About $2.3 billion in private funding |
| OQC | Superconducting computers installed inside data centres | At least $450 million across its two largest disclosed rounds |
| Photonic | Distributed computers using optically connected silicon spin qubits | More than $350 million |
| Pasqal | Neutral-atom computers for analog and gate-based workloads | More than €300 million completed, plus about €170 million tied to its proposed listing |
| Atom Computing | Gate-based neutral-atom computers using ytterbium qubits | More than $300 million in recently announced funding and public support |
| QuEra | Neutral-atom computers moving from analog systems toward fault tolerance | More than $230 million in its latest financing |
| Quantum Motion | Silicon spin-qubit computers made with standard CMOS processes | More than $210 million across its major disclosed rounds |
| Alice & Bob | Superconducting cat qubits designed to reduce error-correction overhead | €180 million |
| Silicon Quantum Computing | Atomically positioned phosphorus qubits in silicon | At least A$70 million across major disclosed rounds and public investment |
| Nord Quantique | Superconducting bosonic qubits with error correction built into the hardware | About $58 million across equity and disclosed government support |
| planqc | Neutral atoms arranged in optical lattices | At least €50 million, excluding additional project contracts |
| Quandela | Photonic computers built around deterministic photon sources | More than €50 million |
| Diraq | Silicon quantum-dot processors made on 300 mm wafers | At least $42 million in equity, with up to $38 million proposed separately under a US funding letter |
Is PsiQuantum really ahead, and who is closest to building a useful quantum computer?
PsiQuantum currently leads the full-system race, while QuEra has shown more fault-tolerance science on working hardware.
PsiQuantum has raised roughly five times as much private capital as QuEra and far more than any other remaining startup. It has started construction work for large quantum-computing sites in Chicago and Australia. Its latest DARPA agreement can provide as much as $125 million to examine and test its proposed machine.
DARPA’s Quantum Benchmarking Initiative assesses whether a quantum computer could create more economic value than it costs to build and operate. PsiQuantum and Microsoft are currently the only two companies in the final validation phase inherited from DARPA’s earlier program.
PsiQuantum is also developing photon detectors, optical switches, chip packaging, fibre connections and semiconductor production. These components must eventually work together inside one large machine.
The company still has no public cloud computer, disclosed quantum-computing revenue or complete machine performing customer workloads.
QuEra has already demonstrated error-corrected algorithms, special quantum states needed for universal computation and repeated correction across large atom arrays. It plans to launch its Libra system through Amazon Braket in 2028, with more than 10,000 physical qubits, 256 logical qubits and capacity for at least one million logical operations.
Atom Computing sits close behind. Its neutral-atom platform contains more than 1,200 physical qubits and has demonstrated logical computation and repeated error correction, although not yet across the full commercial system.
PsiQuantum leads on the complete industrial plan, QuEra on public fault-tolerance evidence and Atom Computing on packaged physical scale.
If you want more recent data on this point, please see our latest quantum computing market report.
As this chart shows, and as featured in our quantum computing market deck, search interest in quantum computing has grown significantly
Who has actually made logical qubits work?
On public logical-qubit evidence, QuEra currently leads the private startup field, with Atom Computing in second place.
A logical qubit spreads information across several physical qubits so the machine can detect and correct errors during a calculation.
QuEra and its research partners first ran error-corrected algorithms using 48 logical qubits built from more than 250 atoms. Later experiments added magic-state distillation, a key ingredient for running a broad range of quantum algorithms.
The group then demonstrated repeated error correction, logical operations, atom-loss detection, replacement of lost atoms and real-time decoding using as many as 448 atoms. Some experiments involved up to 96 logical qubits.
Atom Computing has entangled 24 logical qubits, run an algorithm using as many as 28 logical qubits and demonstrated repeated correction of lost atoms. Its latest work explores toric-code error correction for longer calculations.
Alice & Bob is pursuing lower error-correction overhead rather than a large logical-qubit count. Its Helium system contains an 18-cat-qubit chip designed to produce the company’s first protected logical qubit, but it remains earlier than QuEra and Atom Computing.
Who can actually use and buy a quantum computer today?
Pasqal currently has the strongest combination of installed machines, disclosed revenue and recognizable customers, while QuEra offers the clearest public-cloud access.
Pasqal says seven quantum processors are already in commercial use and another three are in production. Its systems generally contain between 100 and more than 200 neutral-atom qubits and can operate without the deep refrigeration required by superconducting machines.
Aramco has inaugurated a Pasqal computer in Saudi Arabia and opened access through a commercial quantum-computing service. A 140-qubit machine has also arrived at Italy’s CINECA supercomputing centre.
Pasqal reported €16.5 million in commercial revenue for 2025 and more than €66 million in booked and awarded business by early 2026. The larger figure includes grants, so it is broader than customer backlog.
Other named relationships include Crédit Agricole, LG Electronics, CMA CGM, Sumitomo, OVHcloud and Thales.
QuEra has offered its 256-atom Aquila system through Amazon Braket since 2022. Its newer Gemini platform adds digital gates and error-correction testing, with the first system installed beside the ABCI-Q supercomputer at Japan’s AIST.
Atom Computing’s AC1000 offers more than 1,200 fully connected neutral-atom qubits, but access remains selective and the company does not disclose paid usage. Its Microsoft partnership is focused on quantum supercomputers and on-premise logical-qubit systems.
OQC has placed systems in operational data centres. Quantum Motion has delivered a silicon-CMOS machine to the UK National Quantum Computing Centre, while Quandela has installed photonic systems at European research facilities.
None has yet shown a repeatable economic advantage over the best classical computers.
| Startup | Best commercial evidence currently available | What we still do not know |
|---|---|---|
| Pasqal | €16.5 million annual commercial revenue, seven installed processors and three more in production | Customer usage, gross margins and renewal rates |
| OQC | Systems operating in multiple data-centre locations | Revenue, utilization and number of paying users |
| QuEra | Amazon Braket availability, AIST deployment and several government-backed projects | Recurring cloud revenue and hardware sales |
| Atom Computing | AC1000 platform, Microsoft collaboration and external system projects | Pricing, deployment count and revenue |
| Quantum Motion | Complete silicon-CMOS system delivered to the UK NQCC | Whether the system can be sold repeatedly |
| Quandela | Photonic systems delivered to research and supercomputing customers | Current revenue and paid cloud utilization |
If you want more recent data on this point, please see our latest quantum computing market report.
This chart, included in our quantum computing market deck, illustrates yearly VC funding for quantum computing startups
Who has moved fastest lately?
Lately, the strongest overall momentum has belonged to PsiQuantum, while QuEra is advancing fastest in research and Pasqal is building the most visible commercial operation.
PsiQuantum completed a $1 billion funding round, began work at two large computing sites, expanded semiconductor and packaging programs, and signed its latest $125 million DARPA agreement.
QuEra has recently produced research on integrated fault tolerance, high-fidelity neutral-atom gates, faster error correction and lower resource requirements for useful simulations. It has also introduced a clearer 2028 fault-tolerant product plan supported by AWS.
That timeline is later than QuEra’s older plan, which discussed reaching more than 10,000 physical qubits and 100 logical qubits around 2026. The research has strengthened, but the commercial target has moved back.
Pasqal raised another €170 million privately, arranged roughly €170 million of additional financing around a proposed listing, delivered more systems and expanded its commercial team. The listing remains unfinished, so the associated financing is not yet completed capital.
Diraq has made the sharpest jump from a small base. Its latest result with imec demonstrated coherent operation and readout across an eight-qubit silicon array made through a 300 mm CMOS-compatible process, after earlier foundry work focused mainly on one- and two-qubit devices.
Who can build quantum computers at scale and run them practically?
PsiQuantum has the boldest manufacturing plan, Pasqal has the strongest repeat-delivery record, and neutral-atom machines currently look easiest to install.
PsiQuantum designed its architecture around semiconductor manufacturing. Its photonic chips can use established silicon-photonics processes, while separate production lines handle detectors, optical switches, packaging and fibre connections.
The company expects a useful machine to require roughly one million physical qubits. Its main challenge is integrating and operating that many components without losing too many photons.
Pasqal has seven systems in use and three under production. Its machines avoid dilution refrigerators and reportedly consume less than 4 kW, making them easier to install in ordinary computing facilities.
Quantum Motion and Diraq offer the clearest path through conventional chipmaking. Quantum Motion’s first complete machine was fabricated with standard 300 mm CMOS processes and installed at the UK National Quantum Computing Centre. Diraq’s recent eight-qubit array came from imec’s industrial semiconductor platform.
Large-scale silicon systems must still solve wiring, cooling, calibration and control across dense qubit arrays.
Neutral-atom companies avoid chip-fabrication variability because every atom of the same element is identical. Their main engineering problems are laser control, atom movement, loading and reliable parallel operations.
QuEra has operated coherent arrays containing more than 3,000 atoms in research experiments. Atom Computing has packaged more than 1,200 into AC1000, while planqc is developing a 1,000-qubit system for Germany’s Leibniz Supercomputing Centre.
If you want more recent data on this point, please see our latest quantum computing market report.
This chart, included in our quantum computing market deck, looks at IonQ’s strategy in quantum computing
Which big customers and government bets really count?
Government support points most strongly toward PsiQuantum, while customer evidence points toward Pasqal.
Australia’s federal and Queensland governments announced an A$940 million package for PsiQuantum, combining equity, grants and loans. Illinois selected the company as an anchor tenant for its quantum and microelectronics park. The US Department of Commerce has also proposed up to $100 million for manufacturing and component development.
DARPA has taken PsiQuantum further than any private rival by examining whether the complete proposed machine could function and justify its cost.
Pasqal’s customer base is more diverse. Aramco has accepted and inaugurated a machine, Crédit Agricole is extending work into financial applications, and CINECA is integrating a processor with supercomputing infrastructure. LG Electronics, CMA CGM, Sumitomo and OVHcloud add industrial and distribution relationships.
QuEra’s AWS partnership gives it the strongest public-cloud channel in the private market. AWS already hosts Aquila and plans to offer QuEra’s future fault-tolerant system through Amazon Braket.
Atom Computing’s Microsoft relationship combines Atom’s hardware with Microsoft’s software, error-correction work and customer access. The partnership has already produced an on-premise system designed to support as many as 50 logical qubits.
OQC’s work with Digital Realty and NVIDIA gives it a data-centre route, while Photonic has Microsoft as both an investor and technical collaborator.
Whose hardware works best today?
On working hardware, QuEra currently offers the strongest mix of qubit quality, system size and fault-tolerance experiments.
QuEra’s 260-qubit Gemini system reports global single-qubit fidelity around 99.9% and two-qubit fidelity around 99.2%. Those figures sit alongside experiments involving logical operations, repeated error correction and active handling of lost atoms.
A startup can produce a very high-fidelity operation on two carefully selected qubits without showing that the rest of the machine performs equally well. QuEra has tested a broader set of capabilities across larger systems.
Atom Computing leads on the size of a packaged gate-based neutral-atom platform. AC1000 contains more than 1,200 fully connected physical qubits and supports measurement, reset and conditional operations during a circuit. Its strongest published logical demonstrations used smaller sections of earlier hardware.
Diraq and Silicon Quantum Computing have produced excellent results on much smaller silicon devices. Diraq has reported operations above 99% fidelity on foundry-made qubits. SQC reports fidelities reaching 99.99% and a 14-to-15-qubit atomic-silicon platform. These small devices may eventually scale well, but they cannot yet match the system-level evidence from QuEra.
PsiQuantum has published strong component results, including chip-to-chip entanglement and high-quality photonic measurements. Those components have yet to become a complete operating computer.
QuEra leads on balanced system performance, Atom on packaged physical-qubit count, and SQC and Diraq on selected small-device silicon results.
This chart, included in our quantum computing market deck, illustrates yearly funding for quantum computing startups
Which quantum architecture is winning, and whose lead looks hardest to copy?
Neutral atoms are winning the startup race today, while PsiQuantum has built the broadest industrial moat.
QuEra and Atom Computing have demonstrated logical operations across relatively large atom arrays. Pasqal has turned the same broad technology family into installed commercial systems. planqc has won contracts for larger future machines.
Neutral atoms have several natural advantages. Atoms of the same element are identical, they can remain coherent for a long time and lasers can move them into different arrangements. This helps companies build large arrays without manufacturing thousands of slightly different artificial qubits.
The remaining problems concern speed, laser complexity, gate quality across the whole machine and the ability to run many operations at once. Neutral-atom startups have moved furthest experimentally, but they have not removed these bottlenecks.
Photonics has the strongest industrial challenger in PsiQuantum. Light travels easily between modules and photonic components can use semiconductor manufacturing. Photonic and Quandela are exploring smaller or more distributed versions of this idea.
Photon loss remains unforgiving. A system can contain excellent sources, switches and detectors, yet still fail if too many photons disappear while moving through the machine.
PsiQuantum’s moat covers an entire production system: photonic-chip designs, special materials, detectors, optical switches, packaging, fibre attachment, cryogenic engineering, foundry relationships and the ability to finance complete computing sites.
A rival would need to reproduce far more than one successful chip. The same complexity also creates risk because one serious integration problem could slow the whole machine.
Silicon offers the most familiar manufacturing story. Quantum Motion and Diraq can use processes closely related to modern semiconductor production. SQC takes a more specialized approach by positioning individual atoms with extreme precision.
SQC’s manufacturing capability is difficult to copy. It positions phosphorus atoms inside silicon with accuracy measured at roughly 0.13 nanometres and can produce new devices inside its own facility. The trade-off is that Diraq and Quantum Motion may have an easier route to industrial production through conventional foundries.
Superconducting startups benefit from years of engineering progress across the wider industry. OQC uses a more conventional design, while Alice & Bob and Nord Quantique are trying to build qubits that need much less error-correction hardware.
Alice & Bob has concentrated its advantage inside the qubit. Its cat qubits suppress bit-flip errors for unusually long periods, and the company reports a measured bit-flip time above one hour on Helium. Other research groups are exploring similar ideas, so Alice & Bob still needs to turn that head start into larger reliable systems.
Photonic’s moat comes from combining silicon spin qubits with optical links. The company has demonstrated entanglement between separate modules, but it still needs to show a larger operating processor.
QuEra and Atom Computing have strong scientific teams, control systems and accumulated experimental knowledge. Their basic architecture now has several capable competitors, so execution, calibration data, software and error-correction techniques will decide who keeps the lead.
| Architecture | Leading private startups | Why it is competitive today | Biggest remaining problem |
|---|---|---|---|
| Neutral atoms | QuEra, Atom Computing, Pasqal, planqc | Large arrays, flexible connections and the strongest logical experiments | Gate speed, laser control and reliable parallel operation |
| Photonics | PsiQuantum, Photonic, Quandela | Semiconductor production, modularity and easy networking | Photon loss and the integration of huge component counts |
| Silicon spins | Quantum Motion, Diraq, SQC | Very small qubits and access to semiconductor expertise | Cooling, wiring and control across large arrays |
| Superconducting and bosonic | OQC, Alice & Bob, Nord Quantique | Fast operations and a mature engineering ecosystem | Cooling and the number of physical qubits needed for error correction |
If you want more recent data on this point, please see our latest quantum computing market report.
Can we trust the numbers quantum startups publish?
The broad ranking is fairly reliable, while the exact gaps remain much murkier than company presentations suggest.
Peer-reviewed experiments provide the strongest technical evidence. QuEra’s logical-qubit work, Atom Computing’s logical algorithms, Diraq’s foundry results and SQC’s silicon experiments describe their methods and limitations.
Those numbers are still difficult to compare. One company may report its best two-qubit pair, while another publishes an average across the machine. Error detection, correction, post-selection and qubit loss can also be counted differently.
Government evaluation adds another useful layer. DARPA receives technical information that companies do not publish, while installations at AIST, CINECA and the UK National Quantum Computing Centre confirm that complete systems exist outside company laboratories.
Financial claims require the same care. A completed round is real capital. Proposed incentives, letters of intent and financing attached to unfinished listings remain conditional. Commercial revenue should also be separated from broader figures that include grants.
Roadmaps are weaker evidence than experiments. QuEra’s older plan expected a 10,000-qubit, 100-logical-qubit system around 2026, while its current commercial target places the first fault-tolerant product in 2028.
We are most confident about the top four: PsiQuantum leads on the utility-scale plan, QuEra on public fault tolerance, Pasqal on commercialization and Atom Computing on packaged scale combined with logical-computing evidence.
This chart, included in our quantum computing market deck, compares the main business model options for quantum computing hardware startups
Which quantum-computing startups are actually ahead?
PsiQuantum is ahead overall today, with QuEra and Atom Computing as its closest technical challengers and Pasqal leading the commercial market that currently exists.
We place PsiQuantum first because the hardest problem in quantum computing is building a machine that can perform a valuable calculation after accounting for errors, operating costs and construction costs. PsiQuantum has assembled the strongest combination of private capital, manufacturing work, dedicated infrastructure and independent system-level review.
QuEra comes second because it has provided the best public evidence that error-corrected quantum computation can work across a large, reconfigurable machine. It could take the lead if Libra arrives near its current specifications and runs a million reliable logical operations.
Atom Computing ranks third. Its AC1000 platform is already much larger by physical-qubit count than QuEra’s current commercial machine, and its logical-qubit research is serious. A convincing demonstration that uses the full platform for repeated error-corrected algorithms could move Atom into first or second place.
Pasqal takes fourth rather than first because commercial installations still run on pre-fault-tolerant hardware. Yet Pasqal has achieved something its technical rivals have struggled to match: revenue, repeat deliveries and customers willing to host complete machines. Its installation record currently has no close private-startup equivalent.
OQC and Photonic form the next group. OQC is further ahead in deployed infrastructure. Photonic has raised more money and may possess the more scalable architecture, but it has published less complete-processor evidence.
Quantum Motion leads the silicon startups because it has delivered a complete CMOS-made system. Alice & Bob follows because its cat-qubit design could sharply reduce future hardware needs, even though its first protected logical qubit is still being developed.
Diraq is moving up quickly after its recent foundry-made eight-qubit result. SQC has higher small-device performance and a unique manufacturing process, but its bespoke approach may be harder to industrialize. Nord Quantique, planqc and Quandela each have credible technology, although their current evidence is less complete across performance, fault tolerance and commercial scale.
| Rank | Startup | Why it currently holds this position |
|---|---|---|
| 1 | PsiQuantum | The strongest complete plan for a utility-scale fault-tolerant computer, backed by unmatched private funding, manufacturing work and external validation |
| 2 | QuEra | The best public logical-qubit and fault-tolerance evidence, with a credible route toward a cloud-accessible fault-tolerant system |
| 3 | Atom Computing | A very large packaged neutral-atom platform, strong logical-computing research and a deep Microsoft partnership |
| 4 | Pasqal | The clearest revenue and deployment leader, although its current machines remain pre-fault-tolerant |
| 5 | OQC | Strong funding and real data-centre deployments, with limited public evidence on processor scale and paid usage |
| 6 | Photonic | More than $350 million raised and a differentiated distributed design, but less visible complete-hardware progress |
| 7 | Quantum Motion | The strongest complete-system result among silicon startups and a credible standard-CMOS manufacturing path |
| 8 | Alice & Bob | A potentially dramatic reduction in error-correction overhead, with Helium now providing a full system to test the approach |
| 9 | Diraq | Fast recent progress on foundry-made silicon arrays, although the operating processor remains small |
| 10 | Silicon Quantum Computing | Exceptional small-device performance and atomic manufacturing precision, balanced by a specialized production process |
| 11 | Nord Quantique | A promising hardware-efficient error-correction design supported by DARPA and Canadian funding, with limited system-level evidence |
| 12 | planqc | Large neutral-atom contracts and a serious manufacturing program, but weaker logical-qubit evidence than the leading atom companies |
| 13 | Quandela | Real photonic systems and customer installations, although current processors remain smaller and less advanced toward fault tolerance |
If you want more recent data on this point, please see our latest quantum computing market report.
OUR METHODOLOGY
This analysis tests which independent private quantum-computing startup is ahead based on the evidence available today. We compare fault-tolerance research, working hardware, manufacturing progress, commercial deployment, customer adoption, financing, government support and the credibility of each company’s route toward a useful machine.
We include companies trying to build complete quantum computers. We exclude public companies, acquired businesses and startups focused mainly on software, control electronics, sensors, networking or cybersecurity.
We do not treat headline qubit counts as directly comparable across architectures. A physical qubit, logical qubit, analog atom array, photonic component and packaged gate-based processor reveal different things, so each result is judged by what it proves about reliability, control, error correction and system scale.
Peer-reviewed technical work receives the greatest weight. We also use operating hardware, completed financing, confirmed installations, disclosed commercial revenue and independent review by government agencies or customers. Product roadmaps and proposed incentives receive less weight because they describe future execution rather than completed progress.
Funding totals separate confirmed private capital from grants, loans, proposed incentives and financing linked to unfinished public-market transactions. Commercial revenue is also kept separate from broader booked or awarded business that may include grants.
The overall ranking is an aggregation of several different races. PsiQuantum leads on the complete utility-scale industrial plan, QuEra on public logical-qubit and fault-tolerance evidence, Atom Computing on packaged neutral-atom scale combined with logical-computing work, and Pasqal on commercial deployment.
Key sources for PsiQuantum include the company’s development timeline, its technology and manufacturing overview, the $1 billion Series E announcement, the Chicago project announcement, and the construction and DARPA update.
For QuEra, we relied on the Libra and AWS announcement, the current technical roadmap, the peer-reviewed logical-processor experiments, and AWS material covering Aquila on Amazon Braket, the expanded fault-tolerant collaboration, and the Aquila technical overview.
Other central sources include Atom Computing’s AC1000 material, Diraq and imec’s eight-qubit foundry result, the related technical paper, Diraq’s foundry-made fidelity result, Alice & Bob’s cat-qubit stability result, the Helium processor announcement, and the company’s technology overview.
We also used broader peer-reviewed error-correction research, including surface-code error suppression through scaling and real-time quantum error correction beyond break-even, to judge how startup claims fit into the wider technical field.
This chart, featured in our quantum computing market deck, illustrates how revenue is divided among customer segments in the quantum computing market
Related blog posts
- Photonic quantum computing: which startup is ahead?
- Neutral atom quantum computing: which startup is ahead?
